Small molecules that bind specific proteins are powerful tools for controlling biochemical processes, and information on the association of small molecules with proteins can help researchers understand the physical basis of binding, improve computational drug-design methods, suggest the bioactivity mechanisms of other compounds, and, potentially, predict side-effects. BindingDB enables these uses of protein-small molecule association data by providing web-accessible tools for browsing, querying, analyzing and downloading. During the current grant period, we have collected ~40,000 affinity measurements, enriched the web-interface, and linked to other important bioinformatics resources. Usage has grown dramatically. We now propose to take BindingDB to the next level through these aims: Greatly expand BindingDB's core dataset of quantitative binding affinities, establish a new, much broader dataset by automated data collection, and collaborate with new international efforts at cooperative curation and data sharing. Enhance the web interface by creating innovative overviews of relationships among compounds and proteins, new biology- and disease-oriented browsers, and 3D views of molecular structure in the context of affinity. Create new links with other data resources and provide tools for using BindingDB capabilities programmatically. Expand and enhance tools for on-line virtual screening with new methods and links to large compound catalogs. Create educational and tutorial materials. Strengthen BindingDB's hardware and software infrastructure. Respond to community guidance via BindingDB's Scientific Advisory Board, direct user interactions, and a workshop. PUBLIC HEALTH RELEVANCE: This project will expand an NIH-funded, web-accessible database of information about the binding of small, drug-like molecules to disease-related proteins. This will help scientists design new drugs, develop improved computer models for drug design, and use small molecules to probe living systems and learn how they work.